Friday's child may be loving and giving, but Google I/O has already given its all. We've had a day of promising and exciting demos, followed by a day of more pragmatic and incremental improvements. Let's work our way backwards, shall we?

Thursday's Google had far to go to measure up to Day 1. To woo enterprise customers, Google presented the cloud-based Compute Engine, chasing Amazon's AWS Elastic Compute Cloud and, more recently, Microsoft's Windows Azure. The Chrome browser and Drive landed on iOS, leaving Mozilla still coming to grips with Webkit (and, depending on who you ask, falling even farther behind Chrome in overall browser market). Urs Hslzle, senior vice president of technical infrastructure of Google, delivers a keynote on Google Compute Engine

Google also chased themselves by enabling the single most-requested feature of Google Docs: offline editing. Not enough? How about throwing in voice typing as well? Although the online version of Drive provides an increasingly popular alternative to the ubiquitous Word (not to mention a clear winner against the Windows Live version), this addition shows Google recognizing that some of us still need to work when the connection is lost. It also happens to make the Chromebook more of a contender against standalone notebooks.

Wednesday's Google had been full of woe for some competitors.

Android 4.1 has been christened Jelly Bean, the latest in a series of delicious names that help swallow the bitter pill of being a previous-generation Android user (N.B. I'm typing this on a Droid 3). The secret ingredient in Jelly Bean is Project Butter, whose parallel processing of CPU and GPU boosts response times for noticeable gains in performance. Gone is the lag that cursed even the should-have-been-mighty ICS, and the inclusion of triple buffering and Vsync shows that Google is serious about graphics power -- and that they aren't going to take lightly the graphics challenge of the New iPad's highly-touted Retina display (perhaps its main, or only, evolutionary feature).

While staying true to the standard Android design ethic (love it or hate it), Jelly Bean's addition of subtle animations and information-rich notification handling makes the OS more of a polished and professional choice than ever. Other subtle but welcome improvements include better integration between the camera and gallery, and a forward-thinking personal assistant... of sorts.

Google Now is an attempt to take predictive search to the next level, generating "proactive" (ugh) suggestions based on past searches. Add that to a more fluent Voice Search, and you have Google's response to six months of living in Siri's shadow. For all of its forward-thinking diversification, Search is still Google's bread and butter -- after all, nobody will ever say "Appling" or "Microsofting" (at least not in polite company). In practice, however, Now and Voice Search still seem like more of a promising start (in two different directions) than a leap forward to Siri's level.

You have to hand it to Google, they're still trying to turn a minus into a Plus. After last month's photo-intensive upgrade for the Google+ mobile app, there's no doubt that the company is going to keep fighting naysayers. I/O had Google+ trumpeting 250+ million users, launching Events for all of you photo-sharing party planners, and debuting a tablet-specific version of the social network app.

Speaking of which, the Day 1 highlight was the Nexus 7. This small form-factor tablet seems primed to make the most of Jelly Bean, and of the "Pure Android" approach, one that recommended previous direct-sale Nexus devices over their carrier-diluted kindred. With variously believable talk about the upcoming Kindle Fire 2 and 7" iPad, Google managed to be first out of the gate with a cutting-edge device and a welcoming $199 price point.

As a partner, Google chose well. Asus has been the Android underdog to watch over the past year, with highlights such as the compelling Transformer series (no really, Microsoft didn't invent keyboards for tablets). The only conceivable runner-up would have been Samsung, whose relationship with Google is solid, and who has more than proven their mettle in the tablet world. Why these two companies haven't taken more of a bite out of the iPad's market share is still something of a mystery to me, but perhaps a stronger Android OS version will help make up some of the difference.

Google needs a tablet hit. For what it's worth, Microsoft's Surface and Windows 8 are creating a lot of thunder-on-the-horizon excitement typically reserved for upcoming Android devices. Amazon put everything behind the Kindle Fire's holiday season, supporting the plucky little tablet with an impressive app market and a unique focus on e-book qualities. Yet even the power of Bezos could barely sustain consumer interest for a full fiscal quarter. Perhaps eighteen months of base hits by Android tablets are starting to become a monotonous refrain; it's now to the point where nothing but a huge success or TouchPad-level failure can captivate coverage.

For my money (literally), the Nexus 7 has potential. My modest means prevented my tablet desire from finding consummation with either Asus or Samsung, let alone the ice queen that is iPad. And let's face it, the Kindle Fire was a bit of a last-call choice (perhaps not as much as a Coby, but at least I'd allow myself to be seen with the Kindle Fire). However, the Nexus 7 has the right combination of looks and personality, and is a relatively cheap date to boot.

But I am mystified by the Nexus Q. What does it want to be? A paradigm-changing, genre-creating social media server? I like Google Play as much as anyone (except perhaps Katy Perry, who is ALWAYS on the front page), but to base a piece of hardware on Play seems like an ill-advised longshot. For $299 (the same price point that nearly killed the Revue), it's hard to recommend a niche device that is both buggy and needlessly complicated.

I have enough difficulty getting stable DLNA streaming between my Droid and my PC; the Nexus Q compounds this fractional equation by requiring an Android phone, a dedicated app, and an individual Q for each destination device. What would have been nicer? How about a robust media streaming app for the Android phone that people already own? Sadly, I have to concede that hardware restrictions may make this impossible to achieve, at least on the same level as the Nexus Q.

Still, let's face it -- at its best, Android doesn't ever quite achieve the smooth plug-and-play experience of the iOS and iTunes environment, let alone Apple's more obvious $99 video-streaming competitors. And don't forget Roku's near-ideal combination of price and ease of use.

...and then there's those Google Glasses. The April teaser video was a dramatized look at the day-to-day world of the not-too-distant-future, with a side order of modern romance. By comparison, the Google I/O demo showed off the present reality of Glass with in-your-face extreme skydiving applications. The message: Glass is for real, bro... or at least will be in Q1 2012, for developers willing to pre-order a pair for $1500.

Not to be a wet blanket, but all that the demo really showed was that you can mount a webcam to your face. The Google+ Hangout provided a certain amount of videoconferencing cachet, but if that's the current extent of Glass capabilities, we're still light years away from voice integrated Augmented Reality... bro. Sergey Brin, co-founder of Google, wears a Google Glass during a product demonstration.

There are really only two kinds of animals in the Glass menagerie: those who are excited to see what the Glasses can do, and the others who are dismissive and skeptical about Google's vision (or some bizarre animal hybrid like me, who is firmly on both sides). Honestly, both of us would really like a few more details -- such as how the glasses will be powered, and how well the display and voice recognition will actually work. Or even if they can.

Can any battery can provide a usable charge and yet fit inconspicuously on a pair of glasses? If battery tech was truly that advanced, the Galaxy Note wouldn't weigh a pocket-straining 178g, yet get a merely average 10 hours of talk time. On the positive side, Sony Ericsson managed to squeeze out about 12 hours of use (or 3-4 hours of unbroken talk time) from the Experia X10 Mini, which weighs a mere 88g -- but that's still more than ten times the the weight of the average pair of Oakleys.

Even the most capable smartphones (i.e., not the X10 Mini) tend to struggle with voice integration. So far, Siri has clearly been the best of the bunch, but 4S users still find themselves using that virtual keyboard fairly often. And paying four figures to dangle the Android version of Siri in front of your pupils? Probably not too enticing for many consumers.

But I'm prepared to eat my words when the production Glasses start shipping. And I did say "when", not "if"; after all of this extreme hype, Google would face a vaporware embarrassment of Microsoft-like proportions if they decided to consign Glasses to another year or so of development hell.

Output with the old, Input with the new. What do we have to show for it all? We have a very nice little tablet, some kooky glasses, a mystifying media streamer, and the usual suspects of enterprise-courting and OS updates. We're trying hard to keep any WWDC comparisons to ourselves. All-in-all, the usual two-day I/O is looking like a better fit for Google -- although there's more than enough Day 3 sessions to keep developers busy until TGIF time.